Treatment of spinal tuberculosis: conservative or surgical?

Abstracts

Desde a primeira descrição por Percivall Pott da tuberculose de coluna vertebral, tem-se evoluído muito na abordagem dessa doença. Porém ainda existem muitas controvérsias em relação à prevenção das deformidades. Os objetivos de nosso estudo são avaliar as características clínicas e radiológicas de pacientes com tuberculose de coluna e comparar o tratamento conservador isolado com o cirúrgico associado a antibioticoterapia, principalmente quanto à deformidade residual e déficit neurológico. Nosso trabalho é retrospectivo com avaliação de prontuários e radiografias iniciais e ao final do seguimento. O quadro neurológico foi avaliado através da escala da ASIA e as deformidades foram medidas pelo método de Cobb. Foram avaliados 38 pacientes, 11 pacientes foram operados e 27 receberam somente antibioticoterapia. 15 pacientes tinham déficit neurológico, todos melhoraram independente do tipo de tratamento. A média de cifose torácica focal e regional no início do seguimento foi respectivamente 48,8º e 47,86º. Houve aumento da deformidade na região torácica, segmento mais acometido, de 6,3º focal e 9,8 regional ao final de 5 anos. O tipo de tratamento não influenciou na progressão da cifose. Paciente com menos de 15 anos e cifose maior que 30º tiveram pior prognóstico quanto à progressão da deformidade.

Tuberculose; Coluna vertebral; Cifose

ORIGINAL ARTICLE

Treatment of spinal tuberculosis: conservative or surgical?

Rodrigo Serikawa de Medeiros^I; Rodrigo Calil Teles Abdo^I; Fabiano Cortesi de Paula^I; Douglas Kenji Narazaki^II; Leonardo dos Santos Correia^II; Marcelo Poderoso de Araújo^III; Alexandre Fogaça Cristante^IV; Alexandre Sadao Iutaka^IV; Raphael Martus Marcon^V; Reginaldo Perilo Oliveira^VI; Tarcísio Eloy Pessoa de Barros Filho^VII

^IResident Doctor in Orthopaedics and Traumatology, HC/FMUSP

^IITrainee, Spine Group, Orthopaedics and Traumatology Institute, HC/ FMUSP

^IIIPreceptor Doctor, Orthopaedics and Traumatology Institute, HC/ FMUSP

^IVAssistant Doctor, Orthopaedics and Traumatology Institute, HC/ FMUSP

^VPost-graduated Doctor, Department of Orthopaedics and Traumatology, HC/FMUSP

^VIAssistant Doctor, Head of the Spine Group, Orthopaedics and Traumatology Institute, HC/ FMUSP

^VIIChairman of the Department of Orthopaedics and Traumatology, Medical College, USP

^{Correspondences to}

SUMMARY

Much has evolved since Percivall Pott's first description of vertebral tuberculosis. However, there still is much controversy regarding the best approaches to prevent deformities. The objectives of this study were to evaluate the clinical and X-ray characteristics of patients with vertebral tuberculosis and to compare the conservative treatment alone to the surgical one associated with antibiotic therapy, particularly regarding residual deformity and neurological deficit. A retrospective evaluation of baseline and end-point X-ray studies and of the medical files was performed. The neurological status was evaluated by using the ASIA scale and the deformities were measured using the Cobb method. Thirty-eight patients were evaluated: 11 were surgically treated and 27 received only antibiotics. Fifteen patients presenting neurological deficit showed improvement regardless of the treatment method employed. The mean focal and regional thoracic kyphosis at baseline was 48.8º and 47.86º, respectively. An increased incidence of thoracic deformity was found, also being the most affected segment, from 6.3º focal and 9.8º regional after 5 years. The type of treatment has not interfered on kyphosis progression. Patients below the age of 15 and with kyphosis above 30º had worse prognosis regarding deformity progression.

Keywords: Tuberculosis; Spine; Kyphosis

INTRODUCTION

In 1779, Percivall Pott⁽¹⁾, in his monograph, was the first to associate tuberculosis and spinal disease. He described the nosologic and syndromic diagnostic. He noticed that, in patients showing lower limbs’ palsy and spasticity these conditions were caused by an abnormal spine curvature. The onset was insidious, being faster in adults. The overall status showed little changes. The posterior paramedian incision on the torso and drainage of the dense material comprised within vertebrae determined a functional improvement of the affected limbs, allowing patients to ambulate six weeks later. He also noted that the vertebrae showed caries, augmented volume and spongy-like. .

Spinal tuberculosis is the most frequent bone tuberculosis, accounting for 50%. It may be associated to pulmonary diseases or not. If left undiagnosed and not early treated may lead to major sequels such as paraplegia and deformities, which ultimately determine functional loss.

Regarding treatment, literature remains controversial. Compere and Jerome, Chandler and Page, and Cleveland apud Baley et al.⁽²⁾ described their experiences with spinal arthrodesis through posterior approach. Bennett, Fallen and Kaplana apud Baley et al.⁽²⁾ presented their results with conservative treatment. Kidner and Muro, Mayer and Adams apud Baley et al.⁽²⁾ compared the conservative treatment to posterior arthrodesis in children, concluding that no treatment was superior to each other at the acute phase of the disease. However, in the late phases of the disease, when 3 or more vertebrae are collapsed, arthrodesis has shown to be a better approach, determining a lower kyphosis progression.

In surgical approach, we are faced with a new controversy: decompression and anterior, posterior or combined arthrodesis. The posterior fusion was introduced in 1961 by Hibbs and Albee apud Hodgson and Stock⁽³⁾. Decompression plus anterior fusion was introduced by Muller in 1906 apud Hodgson and Stock⁽³⁾ and became popular after Hodgson.⁽³⁾ The advantage of the posterior surgery is that it presents less morbidity, while the anterior surgery can potentially decompress vertebral channel directly, provide injury débridement and fix deformities.

Hodgson and Stock showed that the anterior surgery with autologous rib graft determined a more significant reduction of the disease duration than conservative treatment alone, draining the caseous abscess and allowing for a faster bone fusion.⁽³⁾

Govender and Özdemir^(4,5) showed that homologous grafts from a bone library associated to anterior instrumentation are superior to autologous ribs grafts, presenting a lower rate of graft migration in expense of a later union.

Chen et al. and Klöckner and Valencia R.^(6,7) suggested that for 1-level tuberculosis spondyloscitis, the most effective surgery would be through anterior approach, but when 2 or more levels were affected, the best approach would be a combination of anterior and posterior. Thus, these authors reported lower kyphosis progression.

Currently, conservative treatment is provided with 3 chemotherapy drugs: Rifampicin, Isoniazide and Pirazinamide. The duration of antibiotic therapy is also controversial in literature. Moon et al.⁽⁸⁾concluded, in their study in 2002, that at least 12 months of antibiotic therapy are required. Parthasarathy et al.⁽⁹⁾ conducted a study sponsored by the Chennai Tuberculosis Research Center (Madras) reporting that patients treated with antibiotic therapy alone for 6 months, 9 months or surgery had the same favorable functional evolution (94%, 99% and 90%, respectively), with no statistically significant difference.

The incidence of neurological involvement in patients with spinal tuberculosis is 10 – 46%^(3,10). There are 2 kinds of neurological deficits: early and late. The early deficit occurs during the active phase of the disease; the late deficit, after cure. The previous occurs by direct mechanical pressure (abscess, granulation tissue, debris and caseous tissue), instability, infectious thrombosis and spinal vessels arteritis. The latter, due to spinal kyphosing, where an anterior transverse bone bridge compresses spinal cord⁽¹¹⁾.

Jain⁽¹¹⁾ reports that 30-35% of the patients with neurological deficit recover within 3 to 4 weeks with conservative treatment.

Surgery for those patients with neurological deficit at the acute phase of the disease is limited to cases of conservative treatment failure, fast onset, recurrent paraplegia, uncontrolled spasticity, painful paraplegia, spinal tumor syndrome, and instability confirmed by image. The author recommends the anterior approach for all cases except when the posterior neural arch alone is involved or for spinal tumor syndrome. In the late phase, surgery is recommended in cases of severe kyphosis on thoracolumbar spine.⁽¹¹⁾

In the conservative treatment, deformity can evolve to 15º. Of these patients, 3 to 5% present with a >60º deformity. Deformity progression has two phases: phase 1 – active phase of the disease; phase 2 – after cure. That worsening is influenced by the severity of baseline deformity, injury level and patient’s age. Kyphosis above 30º, thoracolumbar involvement, and age below 15 show higher potential to progress. For each destructed vertebral body, a kyphosing of 30 to 35º exists. Surgery is indicated to avoid this complication⁽¹²⁾.

The objectives of our study are to assess clinical and X-ray features of patients with spinal tuberculosis treated at HC-FMUSP IOT’s Spine Group in the last 25 years and to compare conservative treatment alone to the surgical approach combined with antibiotic therapy, especially regarding residual deformity and neurological deficit.

CASE SERIES AND METHODS

Between 1981 and 2005, one hundred fourteen cases diagnosed with spinal tuberculosis and showing clinical or anatomicopathological evidences were retrospectively studied at HC-FMUSP Orthopaedics and Traumatology Institute.

Cases followed up for less than one year, with incomplete medical data and inappropriate X-ray evaluation were excluded. Therefore, our study included 38 cases.

The neurological assessment was graded according to the criteria of the American Spinal Injury Association (ASIA).

X-ray analysis of deformities was conducted according to Cobb’s method. Focal kyphosis, using preserved terminal plates cranial and caudal to the involved level, and regional kyphosis, using C2 and C7 lower terminal plates for cervical deformities, upper T2 and lower T10 for thoracic deformities, upper T11 and lower L2 for thoracolumbar deformities, and upper L2 and lower L5 for lumbar deformities were measured. Deformities at coronal plane were recorded and measured by using the same method.

In all cases, the following items were assessed: affected vertebral level, presence of early neurological change, symptoms, spinal deformity at diagnosis, co-morbidities and presence of lung tuberculosis, kind and duration of treatment, postoperative deformity in surgically-treated patients, deformity at completion of clinical treatment (for those not operated), after at least 5 years of follow-up, and presence of neurological changes after treatment completion.

The statistical analysis of data was made according to Fisher’s and Mann-Whitney’s tests.

RESULTS

Gender distribution was even, that is, 19 men and 19 women. The age group most commonly affected by the disease was up to 10 years, 50% of patients being below 30 years old. The mean age was 29.92 years, ranging from 1 to 68 years (chart 1).

All patients presented axial pain. Only 3 patients (7%) had sciatic pain. Fourteen patients (36.8%) reported weight loss, and 4 (10.5%), fever. Complaints of deformity were reported by 13 patients (34.2%) and neurological deficit by 15 patients (39.5%). One patient (2.6%) had dorsal fistula, and 2 (5.2%), sphincter changes (chart 2).

We noticed that 6 patients (15.8%) had associated active pulmonary tuberculosis, 2 (5.2%) had kidney tuberculosis, 2 (5.2%) had non-axial bone tuberculosis (1 foot and 1 knee), and 2 (5.2%) were HIV-positive. (chart 3)

The diagnosis of spinal tuberculosis was provided by isolating the agent only in 39.5% of the cases (15 patients).

The most frequently affected level was the thoracic spine, found in 20 patients (52.6%). Thoracolumbar spine in 8 patients (21%), lumbar spine in 9 patients (24%) and cervical spine in 1 patient (2.6%). The most commonly affected vertebra was T9 (11 patients – 29%). At lumbar level, L2 and L4 were the most frequently involved vertebrae (5 patients each – 13%). At cervical level, C1, C2 and C3 were involved.

Regarding the magnitude of the disease, we had 17 cases (44.7%) with 1 level involved, 17 cases (44.7%) with 2 levels affected, and 4 cases (10.5%) with 3 levels involved.

Concerning treatment, 27 patients (71%) received conservative treatment. Of these, 23 patients (60.2%) used some immobilization device (orthosis or cast) for an average time of 4 months (range: 3 months to 1 year).

Eleven patients (29%) were operated on. Of these, 9 patients (23.7%) had thoracic spine injury and 2 (5.2%), lumbar injury. The kinds of surgeries are described on chart 4. (figure 1)

The selection of antibiotic agents widely differed, from simple monotherapy approaches to combinations of four agents. Several antibiotic agents were employed: rifampicin, isoniazide, pirazinamide, ethambutol, streptomycin, among others. The triple approach currently used (rifampicin, isoniazide and pirazinamide) was used in 21 patients (55.2%). The duration of antibiotic therapy was 6 months for 7 patients (18.4%), 9 months for 5 patients (13.2%) and 12 months for 26 patients (68.4%).

Regarding neurological picture, 23 patients (60.5%) had no deficits (ASIA E) and 15 (39.5%) presented neurological deficit. Of these, 11 (73%) had injuries on thoracic or thoracolumbar spine, and 4 (27%) had injuries on lumbar spine. Two patients were ASIA A, 6 patients were ASIA C, and 7 patients were ASIA D.

Among those patients with neurological deficit, all of them presented with progressive improvement after treatment was established, regardless of the kind of treatment, whether conservative or surgical (no statistical difference, p>0.05). Of the 15 patients with neurological deficit, 13 were ASIA E and 2 were ASIA D at the end of follow-up period. The two patients with partial recovery (ASIA D) were those with total paraplegia at diagnosis (ASIA A). (chart 5)

At baseline, we found 21 patients (55.3%) with a mean scoliosis of 8.4º (range: 4º-26º), with no prevalence of side. This deformity remained unchanged until the end of follow-up (> 5 years).

at baseline, we found an average focal kyphosis and thoracic regional kyphosis of 48.8º (20º– 90º) and 47.86º (24º – 90º), respectively. At the end of the follow-up period, an increased level of focal deformity of 6.3º and regional deformity of 9.8º was found on the thoracic region (55.13º and 57.7º, respectively).

We found an average lumbar focal and regional kyphosis of 28.5º(5º – 40º) and 5.5º (-31º – 38º), respectively. At the end of the follow-up period, an increased level of focal kyphosis of 9.25º and regional kyphosis of 19.25º was found on the lumbar region (37.5º and 24.75º, respectively. Please, note that an inversion happened on lumbar physiological lordosis.

We found a mean focal and regional thoracolumbar kyphosis of 22.3º(7º – 30º) and 19º (7º – 30º), respectively. At the end of the follow-up period, an increase was seen on focal thoracolumbar kyphosis of 13.7º and regional thoracolumbar kyphosis of 18.7º (36º and 37.7º, respectively).

For the analysis of the influence of the involved segment on deformity progression, we used only the focal angle, once each segment has its own physiological curve (thoracic - kyphosis, thoracolumbar – neutral, and lumbar – lordosis). We noticed that the level with the worst progression prognosis is the thoracolumbar one (13.7%), this difference being statistically significant (p<0.05). The lumbar level (9.25º) showed more progression than thoracic level (6.3º), but this difference was not statistically significant (p>0,05).

When we compare the thoracic kyphosis progression among patients younger than 15 years (increase of 18.1º focal and 20.4º regional) and older than 15 years (reduction of 3º focal and increase of 2.3º regional), we find that younger patients present a more significant kyphosis progression than older patients do. These data were statistically significant (p<0.05).

When we compare the thoracic kyphosis progression among patients with > 30º deformities (increase of 8.7º focal and 13.3º regional) and with < 30º deformities (increase of 5.3º focal and 7.5º regional), we find that patients with moderate and severe deformities show a more significant kyphosis progression than the ones with mild deformities. The difference of focal progression was not statistically significant (p>0.05), while the difference of regional progression showed statistical significance (p<0.05).

The kind of treatment, either conservative or surgical, has not shown to be statistically significant for thoracic deformity progression (p>0.05). Patients submitted to surgery showed an increased kyphosis by 5.3º focal and 9.6º regional, while for those submitted to conservative treatment the increase was 8.6º focal and 12.1º regional.

DISCUSSION

Over 30 million people are affected by residual tuberculosis (reactivation) and more than 2 million live with the active kind of the disease, its spinal form, worldwide.⁽¹³⁾

Today, this disease can be successfully treated with antibiotics, and our current concerns lie on residual deformity. Severe kyphosis determines a psycho-social and a secondary cardio-respiratory issue to children. Preventing deformities should be currently a priority in spinal tuberculosis treatment. Rajasekaran⁽¹²⁾ in his classical study, reported three radiographic signs that are indicative of spinal instability: facet dislocation, lateral translation, fragments retropulsion and upper vertebra tilt. If a patient presents with at least 2 of the 4 signs, surgery is indicated for preventing deformity progression.

We found that the most commonly affected age group is the one below 10 years old; the most frequently affected level was the thoracic one, particularly the T9 level; the most prevalent symptom is pain in all patients, while neurological deficit was seen in 39.5% of cases, and; associated pulmonary disease was present in 15.8% of the cases. Those data are consistent with literature.^{(2,3,8,9,11-14)}

Regarding neurological deficit, we could see that all patients showed recovery, regardless of the kind of treatment, a fact consistent to the results reported by Parthasarathy et al. and Jain.^(9,11)

Regarding residual deformity, we could see that patients younger than 15, with kyphosis above 30º and involvement of thoracolumbar segment present worse prognostics, being indicated for these cases a more aggressive approach, with anterior and posterior surgery, aiming to avoid its worsening. Rajasekaran^(12,14), in two studies, reports similar data.

Nevertheless, the kind of treatment, either conservative or surgical, did not influence deformity progression, a fact that is consistent to the reports published by Parthasarathy et al.⁽⁹⁾.

This similarity in terms of deformity progression regardless of the kind of treatment could be most likely explained by the fact that we didn’t perform posterior arthrodesis secondary to anterior arthrodesis in our case series, as shown by Chen et al. and Klöckner and Valencia^(6,7), since of the 11 operated cases, 8 were thoracic and 3 were lumbar, and among these, 5 were at multiple levels.

CONCLUSION

Therefore, our recommendations follow the ones by Parthasarathy et al.⁽⁹⁾:

I.Patients with < 30º deformities should receive antibiotics therapy on an outpatient-basis and be followed up by means of X-ray images to detect progression.

II.Antibiotic therapy should be recommended with the triple scheme: Rifampicin, Isoniazide and Pirazinamide for a period of 9-12 months.

III.Surgery is indicated for:

a.Patients younger than 15 years with >30º deformity

b.Deformity progression despite of conservative treatment

c.No neurological improvement despite antibiotic therapy

d.Signs of instability detected by X-ray image

REFERENCES

Received in: 09/28/06, Approved in: 10/10/06

Study conducted at the Orthopaedics and Traumatology Institute, Hospital das Clínicas, Medical College, University of São Paulo.

• 1. Pott P. The chirurgical works of Percivall Pott, FRS, surgeon to St. Bartholomew's Hospital, a new edition, with his last corrections. Clin Orthop Relat Res. 2002; (398): 4-10.
• 2. Bailey HL, Gabriel SM, Hodgson AR, Shin IS. Tuberculosis of the spine in children. 1972 [classical article]. Clin Orthop Relat Res. 2002; (394):4-18.
• 3. Hodgson AR, Stock FE. The Classic: Anterior spinal fusion: a preliminary communication on the radical treatment of Potts disease and Potts paraplegia. 1956. Clin Orthop Relat Res. 2006; (444): 10-5.
• 4. Govender S. The outcome of allografts and anterior instrumental in spinal tuberculosis. Clin Orthop Relat Res. 2002; (398): 60-6.
• 5. Özdemir HM, Us AK, Ögun T. The role of anterior spinal instrumentation and allograft fibula for treatment of pott disease. Spine. 2003; 28:474-9.
• 6. Chen W, Wu C, Jung C, Chen L, Niu C, Lai P. Combined anterior and posterior surgeries in the treatment of spinal tuberculosis spondylitis. Clin Orthop Relat Res. 2002; (398):50-9.
• 7. Klöckner C, Valencia R. Sagittal alignment after anterior debridement and fusion with or without additional posterior instrumentation in the treatment of pyogenic and tuberculous spondylodiscitis. Spine. 2003; 28:1036-42.
• 8. Moon M, Moon Y, Moon J, Kim S, Sun D. Conservative treatment of tuberculosis of the lumbar and lumbo sacral spine. Clin Orthop Relat Res. 2002; (398):40- 9.
• 9. Parthasarathy R, Sriram K, Santha T, Prabhakar R, Somasundaram S, Sivassubramanian S. Short-course chemotherapy for tuberculosis of the spine. J Bone Joint Surg Br. 1999; 81:464-71.
• 10. Griffiths DL. The treatment of spinal tuberculosis. In: McKibbin B, editor. Recent advances in orthopaedics. Edimburgh: Churchill Livinstone; 1979. p. 1-17.
• 11. Jain AK. Treatment of tuberculosis of the spine with neurologic complicants. Clin Orthop Relat Res. 2002; (398):75-84.
• 12. Rajasekaran S. The problem of deformity in spinal tuberculosis. Clin Orthop Relat Res. 2002;(398): 85-92.
• 13. Satyasri S. Global epidemiology of tuberculosis. In: Sathya Sri S, editor. Text book of pulmonary and extra-pulmonary tuberculosis. Madras: Interprint; 1993. p.13-8.
• 14. Rajasekaran S. The natural history of post-tubercular kyphosis in children. Radiological signs which predict late increase in deformity. J Bone Joint Surg Br. 2001; 83:954-62.

Publication Dates

• Publication in this collection
  10 Sept 2007
• Date of issue
  2007

History

• Accepted
  10 Oct 2006
• Received
  28 Sept 2006